The constitutively active androgen receptor splice variants (AR-sVs) lack the LBD and the C-terminus but retain the AR amino terminus domain (NTD), which contains two activation functions, AF-1 and AF-5, the latter of which is critical for their constitutive activity. It has recently reported that the two most common AR-Vs, AR- V7 and ARv567es, direct the expression of a distinct transcriptome that is characterized by genes involved in mitosis and rapid progression through DNA-repair check points. This transcriptome is associated with a more rapid rate of tumor growth and shorter time to death: this has termed this the lethal mitotic phenotype. Collectively, these data strongly implicate AR-Vs or promiscuous AR as the drivers of sustained AR signaling in CRPC leading to more rapid progression and death. Our work on the analysis of the Tumor Cancer Genome Atlas and International Team Stand-Up-to Cancer (SU2C) RNA sequencing data has revealed two new components of the AR-sV program as they relate to prostate cancer. First, we have seen that there is a significant increase in the appearance of AR-sVs in the progression from benign epithelium to primary prostate cancer with up-to 30% of primary prostate cancers exhibiting one or more AR-sVs. When we examined the biopsies from metastases of patients in the SU2C cohort the levels of biopsies expressing AR-sVs was greater than 60%. Of further note in many patients the AR-sVs made up 15-20% of the AR mRNA. This is especially significant when most of these patients were resistant to abiraterone and enzalutamide. These data add significantly to the concept that AR-sVs are a mechanism for resistance to the newest prostate cancer therapies directed at the AR. The second new component relating to AR-sVs that was derived from the RNA- seq analysis was that there were significantly more AR-sVs in the resistant patients than AR-V7 and ARv567es. These variants are described more in the background data of the Research Plan of this proposal. When we made constructs of these variants, we noted that some were constitutively active and others were not nuclear and had no observable effect on AR function. By examining the exon structure of the variants we have gained significant insight into the variant structure necessary to cause therapy resistance. Also, as we will show from our CHiP-sequencing data, the detection of AR-sVs and their cistrome elements may indicate when therapy should be altered in patients. Based on these findings, we propose the Hypothesis: Multiple AR-splice variants drive castration resistant prostate cancer. In order to prove this hypothesis we will undertake three specific aims. In aim 1. we will determine the R-sV repertoire in patient samples of primary and resistant tumors. In addition to detecting the AR-sVs, we will determine their structure. In aim 2. we will determine the components of the AR-splice variants that determine their constitutive activity. Including the mechanisms of transport in and out of the nucleus. In ai 3. we will determine differences and similarities of the variants on gene expression and chromatin interaction. This project will use analysis of previously collected human tissues from various study cohorts, state of the art nuclear transport methods, in vivo xenograft studies to confirm activity of the AR-sVs, and next-generation RNA and DNA analyses. In order to accomplish these goals, we have elicited input from leaders in the fields of RNA analysis, intracellular trafficking of steroid receptors, steroid cistrome analysis, and intratumoral steroid measurement.